Practices as a Unit of Design: An Exploration of ...1043363/FULLTEXT01.pdf · of practice, it proposes that practice-oriented design ... HCI, improvisation actors, performances, practice

21

Practices as a Unit of Design: An Exploration of TheoreticalGuidelines in a Study on Bathing

LENNEKE KUIJER, Delft University of TechnologyANNELISE DE JONG, Interactive InstituteDAAN VAN EIJK, Delft University of Technology

The sustainability challenges facing society today require approaches that look beyond single product-user interactions. Focusing on socially shared practices—e.g. cooking, laundering—has been identified as apromising direction. Building on a growing body of research in sustainable HCI that takes practices as unitof analysis, this article explores what it means to take practices as a unit of design. Drawing on theoriesof practice, it proposes that practice-oriented design approaches should: involve bodily performance, createcrises of routine and generate a variety of performances. These guidelines were integrated into a GenerativeImprov Performances (GIP) approach, entailing a series of performances by improvisation actors with low-fidelity prototypes in a lab environment. The approach was implemented in an empirical study on bathing.Although the empirical example does not deal with common types of interactive technologies, the guidelinesand GIP approach offer sustainable HCI a way to think beyond immediate interactions and to conceptualizechange on a practice level.

Categories and Subject Descriptors: H.5.0. [Information Interfaces and Presentation (e.g., HCI)]:General

General Terms: Design, Theory

Additional Key Words and Phrases: Bathing, design, HCI, improvisation actors, performances, practicetheory, research through design, sustainability

ACM Reference Format:Kuijer, L., De Jong, A., and Van Eijk, D. 2013. Practices as a unit of design: An exploration of theoreticalguidelines in a study on bathing. ACM Trans. Comput.-Hum. Interact. 20, 4, Article 21 (September 2013),22 pages.DOI: http://dx.doi.org/10.1145/2493382

1. INTRODUCTION

When it comes to household resource consumption, apart from a number of exceptions,mainstream research in sustainable HCI focuses on motivating people to “act envi-ronmentally” through eco-feedback or other forms of persuasive interfaces [DiSalvoet al. 2010; Dourish 2010; Froehlich et al. 2010; Woodruff et al. 2010; Brynjarsdottiret al. 2012]. Examples are motivating people to take shorter showers or to turn off thelights when leaving a room. Although persuasive technologies have repeatedly beenshown to render desired effects, it has also become clear that the extent of change, interms of scope and duration, is limited [Gram-Hanssen 2011; Van Dam et al. 2010].

The authors thank the EU FP7 program for financial support.Authors’ addresses: L. Kuijer, Industrial Design Department, Faculty of Industrial Design Engineering, DelftUniversity of Technology; email: [email protected]; A. De Jong, Design Research Unit, Interactive Insti-tute; D. Van Eijk, Industrial Design Department, Faculty of Industrial Design Engineering, Delft Universityof Technology.Permission to make digital or hard copies of part or all of this work for personal or classroom use is grantedwithout fee provided that copies are not made or distributed for profit or commercial advantage and thatcopies show this notice on the first page or initial screen of a display along with the full citation. Copyrights forcomponents of this work owned by others than ACM must be honored. Abstracting with credit is permitted.To copy otherwise, to republish, to post on servers, to redistribute to lists, or to use any component of thiswork in other works requires prior specific permission and/or a fee. Permissions may be requested fromPublications Dept., ACM, Inc., 2 Penn Plaza, Suite 701, New York, NY 10121-0701 USA, fax +1 (212)869-0481, or [email protected]© 2013 ACM 1073-0516/2013/09-ART21 $15.00

DOI: http://dx.doi.org/10.1145/2493382

ACM Transactions on Computer-Human Interaction, Vol. 20, No. 4, Article 21, Publication date: September 2013.

21:2 L. Kuijer et al.

Some authors even argue that feedback devices risk justifying and sustaining energyintensive behaviors. For example, Pierce and Paulos [2010] found that including theenergy use of the automatic clothes dryer in feedback devices communicated it asstandard, thereby inhibiting the switch to the far less resource intensive line drying,a finding confirmed by Strengers [2011].

Inherent in the field of human-computer interaction, these efforts focus on in-teractions. In itself this is practically understandable and not a problem, but incombination with the systemic change required for sustainability [Manzini 2009],this focus becomes limiting. Besides tending to tweak behaviors within existing waysof conduct, interaction-centered approaches are ill equipped to take into accountthe complex and ever changing nature of everyday life [Brynjarsdottir et al. 2012].Therefore, sustainable HCI needs to find ways to look beyond interactions.

1.1. Practices as a Unit of Analysis

Recognizing this need, a number of researchers in product design and HCI are nowdrawing on sociological theories of practice. Theories of practice are relevant for the taskat hand for several reasons. First of all, sociology has a long history of addressing issuesof a societal scale, and theories of practice in particular have proven to be well suitableto understand and explain issues with regard to sustainability (e.g., Røpke [2009],Gram-Hanssen [2009], Shove [2003]). Moreover, because ‘things’ play an explicit rolein certain strands of the theory, practice theory is useful to gain a better understandingof how “designed artefacts shape and are shaped by the contexts in which they are used”[Ingram et al. 2007].

The fundamental unit of analysis in practice theory is practices. Practices are themundane activities that make up most of what people do in their daily lives, such asbathing, cooking, laundering and cleaning, and thus account for the majority of resourceconsumption. Practices are socially shared entities with a certain persistence over timeand space, and are therefore “supraindividual” [Schatzki 2001]. Because a focus onpractices lifts understandings of resource consumption to this supraindividual level,taking practices as a unit of analysis can offer insights into opportunities for systemicchange. As Brynjarsdottir et al. [2012] argue, considering energy in the context ofbroader sociocultural practices rather than focusing on specific, isolated behaviorsmight be more appropriate to the complex nature of sustainability issues.

Researchers in the design field who have drawn on practice theory propose ap-proaches that gain insights from studying particular practices in society, and presentthese insights as a way to inform the design of (interactive) consumer products.Examples of approaches are: studying situated practices [Korkman 2006; Julier 2007],tracing their historic “career” or path [Gram-Hanssen 2008; Hielscher et al. 2008;Munnecke 2007; Pettersen 2009], studying variety in practices between differentcultural groups [Pierce and Paulos 2011; Matsuhashi et al. 2009] and combinationsof these [De Borja et al. 2010; Kuijer and De Jong 2011; Scott et al. 2011]. To giveexamples of design oriented insights that result from these types of studies: Hielscheret al. [2008], from an in-depth study into the history of hair care practices, identifyhow changes in the perception of ‘good’ and ‘bad’ grease have a direct relation to theamounts of water consumed for showering, and Pierce and Paulos [2011], by studyingthe nonmainstream practices of second-hand consumption, reframe technologyconsumption as acquisition, possession, dispossession and reacquisition.

1.2. Practices as a Unit of Design

What unites the approaches listed above is that they are all analytic; they take practicesas a unit of analysis; understanding and gaining inspiration from what is. So far,however, there has been little research on what it means to take practices as a unit


Practices as a Unit of Design: An Exploration of Theoretical Guidelines in a Study on Bathing 21:3

of design; generating and evaluating what could (or should) be in the future. Whatwould it mean, for example, to shift perceptions of grease in hair care, or to reframetechnology consumption for everyday future practices? Therefore, in order to take theopportunities for systemic change further into designed artefacts (products, systems,services), exploration is required on what it means to take practices as a unit of design.

This exploration is relevant for sustainable HCI because it addresses the questionof how to translate opportunities for systemic change into rich future visions thatinform the design of (interactive) systems, while working with the complex socialenvironments—our everyday lives—that they end up in and help shape.

Because this question is cross disciplinary, deals with preferred states, is futureoriented, revolves around a wicked problem and aims to develop theory for design, itpre-eminently lends itself for a research through design approach [Zimmerman et al.2010]. Such an approach leads both to new artefacts, in this case not a product but a“proto-practice,” and at the same time to ideas and knowledge about how to design.

By building on two existing examples of generative approaches that take practices asa unit of design and by introducing an empirical case, this article explores implicationsof practice theory in the generative, future oriented phases of design. To first deducewhat it means theoretically to take practices as a unit of design, it starts by setting outguidelines that a practice-oriented design approach should adhere to. The empiricalstudy is situated in the context of industrial design, however, the article contends andwill elaborate that the methodology presented is equally applicable in the context of(sustainable) HCI design.

2. THEORETICAL GUIDELINES

As a first contribution of this article, literature in sociology and related fields is usedto deduce guidelines or starting points for the “translation” of practice theory, which isdescriptive in nature, to prescriptive, future oriented design methods.

First of all, it is important to stress that there is not one coherent practice theory,but rather a group of similar theories. Therefore, we speak of theories of practice,with main theorists such as Anthony Giddens [1984] and Pierre Bourdieu [1977].Andreas Reckwitz [2002] has provided an overview of similarities and differences ofwork grouped as theories of practice. Theodore Schatzki et al. [2001] have broughttogether, criticized and enhanced these theories. Alan Warde [2005] has introducedtheories of practice into consumer studies, resulting in a lively new field. Finally,sociologist Elizabeth Shove has facilitated introduction of theories of practice into thedesign discipline through a series of workshops and publications in 2005 and 2006[Shove et al. 2007; Ingram et al. 2007]. Thinking about the implications of takingpractices as a unit of design, the article draws on these sources, starting from Reckwitz’swidely used definition of a practice as

a routinized type of behaviour which consists of several elements, interconnected to one other: forms ofbodily activities, forms of mental activities, ‘things’ and their use, a background knowledge in the formof understanding, know-how, states of emotion and motivational knowledge [Reckwitz 2002, p. 41].

2.1. A Constellation of Related Elements

A basic characteristic of practices is that they consist of constellations of interconnectedelements. In his definition, Reckwitz provides a rather loose and nonexhaustive list,making it difficult to work with this nonetheless central concept. Other authors havegrouped the elements in different ways (Gram-Hanssen [2011] provides an overview).The grouping used in this article distinguishes three types of elements: images, skillsand stuff. This is in line with the terminology introduced by Shove and Pantzar [2005]and used in several design-oriented papers [Scott et al. 2011; Kuijer and De Jong 2012].



Fig. 1. Images, skills and stuff model illustrated with bathing example (adapted from Shove et al. [2012]).

Following Reckwitz and Schatzki, this interpretation involves the explicit inclusion ofmaterial elements and therefore clearly speaks to designers. An explanation of thesethree elements below is illustrated with examples from the practice of bathing andsummarized in Figure 1.

Stuff refers to the material elements deployed in the practice. For bathing this is forexample showerheads, taps, soaps, drains and bathrooms. In line with Latour [1993] noclear distinction between humans and things is made; together they can form a hybridentity. For example, a person and a sponge, who can do more than either the sponge orthe person alone. Moreover, as Hielscher et al. [2008] nicely illustrate, “natural things”like skin, hair and grease play determinate roles in the practice of bathing.

Skills are learned bodily and mental routines, including know-how, levels of compe-tence and ways of feeling and doing. Skills involved in bathing are for example knowinghow to judge when hair needs washing, ways of setting the temperature just right, ap-preciation of smells of soap, and so on. Again, there is no clear distinction betweenhumans and things. Skills are distributed, and can be redistributed between peopleand products through what Latour [1992] refers to as delegation. Skills also involveshared knowledge about what is good, normal, acceptable and appropriate, and what isnot. For bathing this includes ideas of cleanliness and public appearance, conventionsof normal bathing frequencies and shared ideas about acceptable levels of smell.

Images are socially shared ideas or concepts associated with the practice that givemeaning to it; reasons to engage in it, and descriptions of what Shove et al. call ‘thesocial and symbolic significance of participation at any one moment’ [2012, p. 22], whichthey explicitly treat as an element of practice, not as something that stands outsideof it as a motivating or driving force. For bathing, images involve getting clean, butalso waking-up, getting warm, getting ready, winding down and having private time,including time to mesmerize and to relax.

Images, skills and stuff together make up practices. Taking practices as a unit ofdesign then means to pay as much attention to novel skills and images, as to stuff.However, this is not all. From their study into the practice of Nordic walking, Shoveand Pantzar conclude that,

new practices consist of new configurations of existing elements or of new elements in conjunction withthose that already exist. From this point of view, innovations in practice are not simply determined by



the generation of new products, images or skills. What really matters is the way in which constituentelements fit together. [2005, P. 61]

In practice-oriented design, the unit of design expands from things to include imagesand skills, but not without consideration of how these fit together.

2.2. A Bodily Performance

To better understand the role of these elements and their links, the distinction Schatzki[1996, p. 89] makes between what Shove et al. [2007, p. 148] summarize as the practice-as-entity and the practice-as-performance is helpful. Practices-as-entity are configura-tions of elements that have a certain endurance over space and time. The practice-as-performance, in which these elements are integrated by people in specific situations,is slightly different each time. The practice-as-entity cannot exist without regular per-formance and it is through the integration of new elements and making new links inperformance that the practice-as-entity changes over time.

Describing a practice-as-entity is necessarily a random indication and a simplifi-cation of something that is in fact highly varied in its manifestations and subject tocontinuous change. However, describing a practice-as-entity, as opposed to describingone single performance, is more generally recognizable and in a way enables capturinga larger number of situated performances. This does not mean that the practice-as-performance can be cast aside. Drawing on theories of practice, performance is centralin the conceptualization of change in practices, and when speaking about performance,the human body receives particular attention.

In practice theory, the human body plays an explicit role. In the words of Reckwitz[2002] practices are “routinized bodily activities.” Or as Schatzki puts it [2001, p. 3].“The skilled body commands attention in practice theory as the common meetingpoint of mind and activity and of individual activity and society.” This attention isrequired for two reasons. One, because without involving the body, an important realmof knowledge is neglected and two, because innovation in practice happens throughactive performance.

Through performance, the body becomes trained in a certain way, when knowl-edge about the practice becomes embodied in the practitioner. In other words,people(‘s bodies/minds) are carriers of practices [Reckwitz 2002]. Practice theory rec-ognizes that there are things our body knows, but that we cannot verbally express. InGiddens’ Theory of Structuration, this knowledge is present as practical consciousness:

practical consciousness consists of all the things actors know tacitly about how to ‘go on’ in the contextsof social life without being able to give them direct discursive expression [Giddens 1984, p. xxiii]

For example, we can all walk, but cannot exactly put into words how we do it. Withoutbodily performance as part of the design process, an important realm of knowledgewould be left out of the equation.

Moreover, when generating novel constellations of elements, bodily performanceis important for another reason. People are carriers of practices, but not passively.Practices-as-entity are maintained and reproduced through regular performance[Shove et al. 2007, p. 19], but through varieties in performance, they are also trans-formed. When taking practices as a unit of design, bodily performance (i.e., activeintegration) should take a central position, for

innovations in practice depend upon the active integration of elements, some new, some already wellestablished, that together constitute what we might think of as innovations-in-waiting or proto-practices[Shove and Pantzar 2005, p. 48] (emphasis added)



A proto-practice, in which familiar and novel elements are integrated, emerges fromactive bodily performance. Such proto-practices could be the result of practice-orienteddesign activities. However, in practice theory there is never an end result, becausepractices constantly change. In fact, change is default.

2.3. Crises of Routine

When zooming in on practices-as-performance, it becomes clear that this is not a uni-form or constant picture. The practice-as-entity forms a collection of elements andpotential links that allow for a wide variety of performances. However, a performancecan also integrate elements that are not part of the practice-as-entity or makes linksthat have not been made before. In such cases, routine performance is not possible andpeople will “adapt, improvise and experiment” [Warde 2005, p. 12].

These instances of adaptation, improvisation and experimentation in performancecan be triggered by all kinds of smaller and larger changes in circumstances, such asfor example the introduction of unfamiliar elements. Reckwitz describes such circum-stances as “crises of routine” [2002, p. 255] in which the “breaking” and “shifting” ofstructures takes place. Through these particular types of “crisis” performances, the ex-isting practice is reconfigured into novel variations that involve both new and existingelements, and new and existing links. To generate desirable reconfigurations of thepractice, such crises of routine could be deliberately staged.

2.4. Variety of Performances

In itself however, an improvised performance does not change the practice-as-entity.Only if this variety on the practice is repeatedly performed by several practitioners (if itworks and spreads), it can change from exceptional and improvisational to acceptableand normal, thereby reconfiguring the practice-as-entity.

Simply because each performance is different, a similar crisis will, through varied(improvised) performances result in a range of reconfigurations. In each performance,different elements are available and different links seem to make sense. Given thecrisis, a common reconfiguration that works emerges from a variety of performances.Therefore, taking practices as a unit of design implies generating a wide variety ofpossible performances.

Summing up the previous, it can be concluded that taking practices as a unit ofdesign means to generate reconfigurations of images, skills and stuff. To achieve this,an approach should involve bodily performances, create crises of routine and generatea variety of performances.

3. EARLIER EXPLORATIONS

Published examples in which practice theory is explicitly used in the generative phasesof a design project are few, but the two that are there form valuable explorations to drawlessons from. One is the “Experiments in Practice” approach developed by Scott andcolleagues, the other is the “Trigger-Products” approach developed by the authors. Bothwill be shortly described and reflected upon in the light of the guidelines introducedabove.

3.1. Experiments in Practice

Scott et al. [2009] have developed a multifaceted approach in which theories of practiceare integrated with principles of codesign. Focal generative element in the approach isExperiments in Practice, in which participants are challenged to conceive and performpractical experiments in the context of their own homes and daily lives. Goal of theexperiments in their pilot study was to try and make the mundane activity of bathing“more effective or sustainable.” During the two-week study, participants were guided



by a workbook and interacted on a closed blog. They met twice, both at the start andthe end of the two weeks.

The idea of experiments clearly involves bodily performance and thereby theintegration of images, skills and stuff. Moreover, the task to experiment created crisesof routine in which participants were challenged to explore and stretch the borders ofnormal practice, thereby creating novel configurations of elements. The combination ofinvolving a group of participants, and the duration of two weeks resulted in a varietyof performances.

However, as the authors already mention, coming up with alternative configura-tions is not easy; “it requires creativity, courage, and the ability to envision somethingentirely new” [Scott et al. 2011, p. 7]. The authors further observed that creativityin challenging normal practice was complicated by the “rigidity of existing bathroomstructures.” As a consequence, most of the experiments largely stuck to the existingpractice of daily showering, for example by reducing the use of cosmetics or shorteningshower time.

3.2. Trigger-Products

The idea that new things can form leverage points or triggers for playing out moreradically different configurations, guided the setup of the Trigger-Product study. Inthis study on thermal comfort, prior analysis pointed in the direction of person heatingto complement space heating, thereby widening the variety of ways to achieve comfort.Earlier studies suggest that offering such variety can reduce dependency on mechanicalheating and cooling and thus energy consumption in households [Kuijer and De Jong2012].

Participants in 60 households were offered a simple, small heat source and asked “totry out the product and see how you can use it as a way to keep warm,” to “be creativewith the product,” and to report on how they used and experienced it in the context oftheir own homes and daily lives. The trigger-product was not selected because it was initself thought to be a “sustainable innovation” or a “solution” to achieve reduced energyconsumption for indoor heating. Rather, the goal of the study was “to explore what typesof uses would emerge and how these may fit in or conflict with existing practices,” andto trigger “a variety of bodily responses and the evaluative reflections they evoke.”Because person heating is currently not widely practiced in the Netherlands, the aimwas to get a preview of how such practices might be composed and find a place inexisting complexes of domestic practices.

From the study it became clear that quick and local heat could work in situationsof passivity such as watching TV or reading a book. Conflicts arose when it turnedout that such heat sources require more work and attention than the passive relationpeople are used to having with their indoor climate control. Also found were undesirabledirections of development, such as taking the heat source outside, potentially leadingto increased expectations of comfort outdoors.

Again, bodily performance and experimentation is asked from the participants, inwhich integration of existing and novel elements (the trigger-product, existing expec-tations of required work) took place. The exercise led to conflicts and misfits in dailypractice, which represent crises of routine. The large number of participating house-holds led to a wide variety of manifestations of person heating practices.

However, the authors concluded that “exploring how person heating may work incurrent household situations, with existing skills and images, might not be represen-tative for the future situations that such a product may end up in and help shape.” Inother words, the trigger-products were used in a context of existing skills and images,which were difficult to let go of in the setting of a normal domestic situation.



3.3. Core Elements and Limitations

Summing up, the approaches both entail recruited participants—“ordinary people”—coming up with and trying out different ways of going about very mundane daily things,such as in these cases bathing and staying warm, in their own homes. Both examplesinvolve bodily performance, crises of routine and generate a variety of performances.However, they also reveal some limitations of this type of setup.

Although leading to valuable insights on potential alternative ways of bathing andstaying warm, and product design opportunities and challenges associated with them,it became clear that participants tended to generate reconfigurations that largely stuckto existing (resource intensive) practices. A challenge is therefore to find a set-up thatis able to break with the rigidity of existing structures in existing household situationsand to find participants with the required creativity, courage, and ability to envisionsomething strongly different.

4. GENERATIVE IMPROV PERFORMANCES

The empirical study presented here followed a research through design process andthereby addressed both empirical and methodological aims. The methodological aimwas to incorporate the guidelines of taking practices as a unit of design into a generativestudy, thereby explicitly looking for ways to strongly reconfigure existing practices.The empirical aims were situated in an ongoing project on the topic of bathing—as awashing or soaking (as in water or steam) of all or part of the body [Merriam Webster2012]—that had been running for several years.

4.1. The Bathing Studies

Water and energy use for bathing take up a large share of household direct resourceconsumption, and have grown sharply over the past decades. Statistics on water use forbathing from before 1970 are difficult to find, but from knowledge that proper body carein the 1950s meant daily washing of hands and face and a (shared) bath once a week[Dutch open air museum 2012; Overbeeke 2001], average weekly water use per personwill not have exceeded 50 liters. By the 1970s, the separate bathroom with shower hadbecome a more common occurrence in Dutch homes and the shower started to overtakethe bath as a popular way of full body washing. Today, showers are standard in everyhome and showering has become the dominant means for full body washing.

The shift from the flannel wash and bath combination to showering went hand inhand with an increase in water consumption. In just sixty years, water use for bathinghas increased from under 50 liters per person per week, to over 50 liters per day.The shower, with its convenience, played a part in increased frequencies of full bodywashing, which went from once a week in 1950, to 5–6 times per week in 2010. Also, withits continuous flow it contributed to an increase in consumption per bath: an averageDutch shower requires 67 liters [Foekema and Van Thiel 2011]. Closely related, Handet al. [2005] explain, were society wide changes in perceptions of the body and the self.While, cleanliness has been a prerequisite for participation in normal social life sincethe late 19th century [Bushman and Bushman 1988], standards of cleanliness slowlyincreased over time. Moreover, showering, with its qualities of “invigoration, freshnessand fitness” [Hand et al. 2005] closely reverberates late-modern ideas of “the body as avehicle of pleasure and self-expression” [Featherstone 1991: 170 in Hand et al. 2005].

Taking notice of how technologies, ways of washing and standards of proper bodycare are closely related, to each other and to levels of water consumption, it is clearthat bathing is subject to change. Combining these insights with studies into bathingin other cultures, where for example in India a bucket bath requires less than 20 liters[Matsuhashi et al. 2009], it can be concluded that (close to) daily showering, with its



Table I. Relation between Elements of GIP and Guidelines for Taking Practices as a Unit of Design

Involve bodily Create crises of Generate variety ofperformance routine performances

The stage Space to move and thingsto act with

Remove essential element andintroduce unfamiliar stuff inform of prototype

Underdeterminedlow-fidelity prototype,variety of props

The scene Task to perform in detailfrom beginning to end

Challenge to act as normal insituation that isn’t, restrictionon the use flowing water

Unscripted scene, requireimprovisation

The players Trained to use their bodyin creative ways

Skilled to improvise, say ‘yes,and’ instead of ‘no, but’ in face ofa challenge

Variety of people inmultiple performances,people skilled andmotivated to act out ofthe ordinary

constant flow of warm water is not the only and possibly not the least resource intensiveconfiguration.

Starting out with a general aim of exploring how practices of bathing could becomeless resource intensive, the objective has been refined during the project. Having anal-ysed the practice of bathing as it is, the focus of the project shifted to questions ofwhat less resource intensive forms of bathing might be in the future, that is, bathingas a unit of design. With this renewed focus, a concept was selected that was broadlydefined as “a way of bathing that is based on contained rather than flowing water.” Itwas coined with the working name splashing [Kuijer and De Jong 2011].

The main empirical question became whether and how splashing could work as aless resource intensive reconfiguration of existing bathing practices in the Netherlands.This empirical question was divided up into three subquestions: (1) what could splash-ing be as a novel bathing practice: what are dimensions of variety in performances andwhat practice-as-entity emerges from them, (2) what could be potential future effectsof a shift from showering to splashing on household resource consumption and beyond,and (3) what are requirements and insights that can inform further refinement of theconcept of splashing.

The study involved a lab setting of a simulated bathroom with a rough prototype inwhich improvisation actors are asked to perform a fictive scene of their splashing “rou-tine” in detail, followed by an interview. The rationale behind this set-up is explainedin terms of the stage, the scene and the players, and is grounded in the theoreticalguidelines. An overview of the relation between the set-up and the guidelines is offeredin Table I.

4.2. The Stage

To enable a bodily performance, some form of physical environment was required. Forthis study, the choice was made to take people out of their everyday context into alab environment. This choice was based on the observation, in previous studies, thatparticipants found it difficult to let go of existing routines (skills) and norms (images)in the fixed environment of their bathrooms and shower booths (stuff ). A lab, like astage, was expected to form a place where acting out of the ordinary is more normalthan at home.

In the lab, a bathroom-like space of approximately nine square meters in size wasequipped with different props like a bath carpet, a bathrobe, soaps, plastic ducks,towels and sponges (Figure 2). Instead of the familiar bathing fixtures (bath, sink,shower) there was a rough, low-fidelity prototype consisting of a 20-liter basin on anintegrated stand, and a seat. The prototype was deliberately left open; there were for



Fig. 2. A selection of props available in the simulated bathroom and the rough prototype designed and madeby Harish Karakat [2009].

example no buttons, taps or drains on it and was made of foam with uneven plasterand cardboard parts attached with sellotape (Figure 2).

Such a low-fidelity prototype in a lab environment however, was not suitable to beused with water. Next to raising practical issues, adding water would have requireddecisions on where to place a tap, faucets and drains and how to shape them, thusmaking the prototype more determined. Therefore, the choice was made to, in thisstage of development, have a dry bathroom with only imaginary water. This allowedparticipants to participate wearing their normal clothes, something that greatly helpedrecruitment.

For capturing the performances, four cameras and a microphone were installed. Theimages were broadcasted live for the researcher, but out of sight of the performer inorder to create a semi-private bathing situation.

4.3. The Scene

The physical environment provided an open, yet subtly steering stage for a bodilyperformance. To further create a setting from which novel links between stuff, skillsand images could emerge, a fictive situation was created in the form of a rough sceneoutline. The instructions in the study mimicked the types of instructions routinely usedin improvisation acting, in which Medler and Magerko [2010] distinguish functionalconstraints and content constraints. A summary of the instructions is as follows.

—Imagine this is your own bathroom and that splashing is your normal way of bathingwith which you are satisfied.

—Perform a complete splashing session starting by entering the bathroom, at leastwashing your body and hair and leaving the bathroom after finishing.

—Pretend to be bathing and make a credible performance with eye for detail, payingspecial attention to the water, you can imagine the space to be comfortably warm.

—You can use all objects you see around you and if you want to use something that isnot there, imagine it; the only thing you cannot use is a shower hose or shower headwith continuously flowing water.

Removing the shower from the bathroom, effectively destabilizes the normal practiceof showering resulting in a serious crises of routine. At the same time, the setting offers



some alternative elements in the form of the rough prototype, which is presented aspart of a way of washing that works. By sketching a scene outline in which splashingis considered routine and accepted, the setting challenges participants to dig deep intotheir repertoires of images and skills to create, through performance, a reconfigurationthat works and is strongly different from what is normal today.

After the performance, interviews were held with the participants while still in theirrole. It focused on eliciting the explanations that are used to position splashing as anacceptable form of bathing. The interview mimicked the type of casual conversationin which one person tells another about a new experience. This is based on Giddens’sconcept of establishing (in this case also generating) mutual knowledge and validdescriptions of social activity [Giddens 1984, p. 375].

4.4. The Players

Future products can be simulated with prototypes, but how to simulate future practi-tioners? Because practices and with them society constantly change, future practition-ers are carriers of practices that contain different things, knowledge, know-how, andideas of what is good and normal. While being provided with a challenging situation,participants needed to have the skills required to make it work. For this study, theinvolvement of improvisation actors was sought, with the expectation that they are:

—trained to improvise and thus make situations work by coming up with creative waysof doing,

—used to working at and over the borders of what is considered normal and appropri-ate, because “actors free themselves from socially accepted frames of reference andassumptions of expected behavior” [Vera and Crossan 2004];

—used to imagining things that are not there, such as water.

Moreover, for this study specifically it is useful to work with people who are nothesitant to perform a private activity, like bathing in detail while being observed.Naturally, these participants were at the same time experienced practitioners in thepractice of showering.

To make the distinction clear, the setup is not improvisation theatre. The studymakes use of the specific skills of improv actors and uses the functional and contentconstraints typical in improvisation theatre, but there is no audience, no objective toentertain and no competition element.

Table I summarizes the rationale behind the setup in relation to the theoreticalguidelines introduced earlier. For reasons of convenience, this set-up is further referredto as Generative Improv Performances or GIP.

5. PERFORMING SPLASHING

Eventually, 17 participants took part in the study, nine of which were trained impro-visation actors. There was an equal spread in men and women and a wide spread inages, from under ten to just over fifty. Most participants were Dutch, but other culturalbackgrounds were Mexico, Greece, and the Philippines. The data of the study con-sisted of video (of the performances), audio (of the interviews) and notes made by theresearcher. Together, the 17 participants produced a total of 25 performances of splash-ing; most participants performed twice, either in a duo performance as requested bythe researcher (3 cases involving 6 participants), or in a second individual performanceon their own initiative (5 cases). All performances were videotaped. The videos haddurations of 00:2:10 minutes to 00:19:46 minutes, summing up to a total of 4 hours.The audio summed up to a total of 2.5 hours. Both audio and video recordings weretranscribed; the videos in terms of both the actions and utterances of the participants.Data was analyzed using qualitative data analysis software in which utterances and



Fig. 3. Graphical overviews of the performances of Teun and Anneke. c© Lenneke Kuijer.

actions were tagged and clustered. When names of participants are used, they arealiases.

5.1. Splashing as a Proto-Practice

The term showering comes with a relatively coherent set of collective, general knowl-edge, know-how, things and meanings connected to it. This is because showering existsas a fairly common practice. The term splashing does not have such collectively sharedluggage. In order to evaluate potential future effects of splashing, on warm water con-sumption and beyond, and to inform further development of the concept, splashingneeded to be fleshed out. Therefore, one of the aims of the study was to generate a richnarrative of splashing through a variety of performances, and to distil from these apreview of splashing as a relatively coherent reconfiguration of elements.

First of all, the 25 individual performances were analysed in detail and summarizedin graphical overviews. These overviews use a pictogram library, which itself emergedfrom analysis of the performances (see Figure 3 and Figure 4).

Next, the different aspects of variety resulting from the set of performances weresummarized (Table II). The table is not exhaustive, but rather gives an impressionof the types and ranges of variety found. Dimensions of variety were sought in allelements of the practice and in the relations between them, with specific attentionfor warm water consumption. For example, “ways of wetting” mainly involve skills,“interactions” zooms in on the stuff in relation to skills, “likes” and “dislikes” surface



Fig. 4. Part of the pictogram library. c© Lenneke Kuijer.

aspects related to image, and “durations” and “water use” give estimations of resourceconsumption.

From this variety of performances, a preview of the practice-as-entity emerges. Re-sults of the study indicate that splashing could be an active, flexible way of washingthe body, with water from a basin, involving sitting and standing postures, a rangeof ways of applying water and soap, involving scoops, sponges and hands, in varyingsequences. Rather than rinsing with constantly flowing water, soap plays a central rolein cleaning the body. Splashing can be quick and functional, washing selective parts ofthe body, but also a relaxing, time taking ritual with a focus on scent and deliberatebody care.

Such a description may not mean that much to nonpractitioners. It is thereforeinsightful to mirror splashing with its direct benchmark, showering. In such a view,showering requires large amounts of water, can be tiring because you have to standall the time making it difficult to wash your feet, is inflexible because you have to wetyour entire body without being able to direct what is wet or rinsed when, and soap isrinsed off before you’ve had time to appreciate it. Water is splattering everywhere andthere is a lot of steam so you cannot keep a towel or clothes nearby, and your bathroomgets humid. In other words, splashing repositions showering.

Having generated a preview of what splashing could be like beyond the abstract:“a way of washing that is based on contained water,” it is now possible to go into moredetail on potential future effects of a shift from showering to splashing.

5.2. Potential Future Effects

When looking at the (virtual) amounts of water participants required for splashing, thestudy indicates that it can be considerably lower than showering. Important reason forthis potential lies in the decoupling of water use from bathing duration; in splashing,taking time does not necessarily mean using more water. For example, to wash herhair and body, Anneke required a total of 4.5 liters (15 scoops of 300ml), taking around



Table II. Overview of Variety in Performances

Washing sequences Head to feet, feet to head, from torso outwards, only upperbody, start with hair, end with hair, intermediate dressingand drying, dressing and drying at end, soap hair rinsehair then start with body, soap hair soap body then rinsebody and hair

Ways of wetting Scoop and pour or splash with cup or bowl, soak andsqueeze sponge or wash cloth, splash with hands, immersebody parts in basin (arm, foot, head)

Ways of soaping Make soapy water in basin and rub with sponge (with orwithout separate wetting first), put soap on hand and rubon body, put soap on sponge and rub on body

Ways of rinsing Scoop and pour, rub with washcloth then rinse and wringout, splash and rub with hands, with towel together withdrying, immerse hand or foot in basin and splash withhands

Actions besideswashing

Shaving, drinking, brushing teeth, listening to music,applying body lotion and make-up, playing, relaxing

Postures Sit on seat, stand in front of basin, put foot on seat, putfoot in basin, sit on basin, bend over basin, lean on basin

Durations From 2 minutes to 19 minutesWater use From approximately 4 liters of actual water for washing to

over 100 liters (several refills basin, flooded floor, tap anddrain open continuously, water jets from basin and walls)

Interactions Place things like sponges, soap bottle, washcloth, razor inor on edge basin and seat, have tap above basin, on edgebasin, integrated in sides or bottom basin, on wall, onseat, with digital display, with turn knobs, with foot pedal,by rubbing side basin, have mirror above basin

Likes Saves water, it’s quick, you can take your time, it’srelaxing, it is comfortable, it is deliberate, enjoy thefeeling, it stimulates the senses, it is flexible, it is cosy, it isfun, it is compact, it creates little moist

Dislikes Do not feel satisfactorily clean, makes a water mess, it isslow, it is not relaxing, it is boring, awkward to sit naked,miss the flowing water

10 minutes and Maartje used about 50 liters in 19 minutes. In comparison, a 19 minuteshower, even with a water saving showerhead, would take 141 liters of warm water.However, this does not mean that splashing cannot be quick. Marcel used as little as4 liters in 4 and a half minutes, and Magnus needed just 2 minutes for a quick wash,using approximately 10 liters.

It has to be mentioned that these amounts are estimations made on the basis ofperformances that did not include actual water, nor the experience of wet nakednessand vulnerability to cold that come with it. The exact numbers therefore mean little.What matters is their order of magnitude in relation to the bathing strategies theyentailed. In addition to the inherent reliance on contained water, such strategies forexample included combining wetting and soaping by making soapy water in the basin,and washing the body in stages (e.g. washing the hair while wearing a bathrobe).

As in the Trigger-Product study, potential negative effects of a shift from showering tosplashing emerged as well. For example, a shift of focus to soap and a link between soapsand pleasure could pose future challenges. It is not clear at this point whether splashingwould require more or less soap than showering—soap is rinsed away less, but becomesmore central to the practice—but the use of more soap would mean a negative effect forthe environment in terms of water pollution and resource consumption. It also raisesquestions about how splashing would work for people with sensitive skin or those



currently not using soap for bathing. Furthermore, splashing, being so flexible, mightincrease bathing frequencies, having for example a quick session in the morning and alonger one in the evening.

It is clear that the potential effects of splashing differ in scale from earlier studies inHCI that have focused on resource consumption in bathing (e.g., Kappel and Grechenig[2009], Ravandi et al. [2009], Laschke et al. [2011]). These studies focus on motivatingor persuading people to shorten their shower duration through feedback and persuasivetechnologies, such as a pattern of colorful dots or a cord with LED’s. Not only is theorder of magnitude of possible savings of a shift to splashing larger, the chances ofenduring change might be higher. Different from a feedback device that can simplybe removed or ignored, splashing, if it would spread, would be anchored in place by adifferent fabric of the bathroom, a different set of bodily expectations and related to analtered set of meanings.

Naturally, such a shift is more drastic than placing a timer in your shower, which arealready in the market. Possibly however, these devices to motivate people to showershorter could facilitate a shift to splashing. Because they problematize showering asa water intensive practice, the disjunction [Hand et al. 2005] that may arise betweenthe spread of power showers and images of wellness in the bathroom on the one hand[ISH 2011], and ideas of ‘long’ showering as problematic on the other, may lead toopportunities that splashing could take advantage of.

Although the current study has given some insight into potential future effects ofsplashing, all of the points above are in need of further study, the starting point ofwhich will be a redesign of splashing.

5.3. Refining Splashing

The third empirical objective of the study was to generate recommendations for furtheriterations of the splash concept. With practices as the unit of design, skills and imagesbecome part of what is “designed.” This is different from approaches where productsare designed to fulfil what are taken to be static needs, as in common user research[Shove et al. 2007] or match existing skills as in most forms of user-centered design[Scott et al. 2009]. Instead of trying to “solve” possible dislikes and “usability problems”with technology, these issues are critically deconstructed, while skills and images areacknowledged as subject to change.

That some participants found splashing slow, not relaxing or boring, where othersfind it quick, relaxing or fun does not mean that splashing is these things. Rather, thisshows that all these different links were made through the performances and thus thattheir potential is there in the current design. Further development means channelingthe development of these links in beneficial directions. For example, the deliberate wayof washing the body that is part of splashing was considered as a challenge,

In showering I had the feeling that everything you forgot would at some point be rinsed off, also soapresidues and possibly a spot you had not reached would become wet anyway and in splashing you have torather explicitly think about all parts of your body. So in that sense you need to be more attentive. (Ralf )

or quality by different participants.

you have to start up more calmly (than when taking a shower) because you are going about things moreconsciously, like, it sounds maybe crazy, ‘yes, now I will wash my legs’ or something, but that’s it, youhave to be more relaxed, because if you do it quick and roughly you will not feel clean, but if you takethe time to really scrub, well, on each limb then yes, nice, just peacefully start up, yes, that is a greatadvantage, yes, yes. (Maartje)

Making sure that all body parts are washed and rinsed properly can be viewed assomething of a hassle inherent to splashing. Or, it can be something that is learned as



a matter of course if deliberately addressing all different parts of the body is positionedas a way to relax.

Another option of dealing with dislikes is allowing for variety. The issues someparticipants had with sitting down naked are not a major problem if splashing allowsboth sitting and standing postures. Moreover, because some participants did not mindsitting, and sitting naked is done routinely in other cultures and (home) sauna’s, itclearly seems like a bodily skill that people can learn. It might very well be that thisissue will disappear spontaneously over time.

The issue of missing flowing water is a more tricky one that touches the core ofthe concept. In spite of the explicit instruction not to use a shower with continuouslyflowing water, participants found ways to incorporate flowing water in their routine,for example by having a tap above the basin that they left open to rinse their hair, upto several jets that were oozing water from the edges of the basin into the bathroom.However, from the observation that most of the participants, 14 out of 17, did notexplicitly mention missing flowing water, it seems that splashing can offer qualitiesthat can make up for this absence. Further study and development of the concept hasto show whether this is indeed the case.

Besides rendering insights on skills and images, the study also resulted in adjust-ment of the brief for the stuff of splashing. For example, in terms of water use persplash it became clear that 20 liters is too large for the basin. The estimations of wateruse offered above have been made by summing up amounts of water applied to thebody through scooping, sponging or splashing with hands. When looking at the wateractually used in the process, this is quite a bit more. Reason for this difference is thatwhile most of the time filling the basin fully, participants used much less than the20 liters content before they drained again. For example, for washing his hair, Teunused a total of fourteen 300ml scoops of water. He then drained and refilled the basinto wash his body, using 17 scoops and drained again. While drawing off 40 liters, heactually used 9.3 liters of (virtual) water to wash. Moreover, the time the participantsestimated for the basin to fill was much lower than the about 2 minutes it wouldrealistically take. This observation, combined with remarks that quickly filling basinis part of their imagined ritual, supports the conclusion that the basin should be madesmaller.

A second design requirement that emerged was to explicitly position splashing in awet space. This could address the dislikes of not feeling clean and creating a water messat the same time. Feeling satisfactorily clean was hampered by different reasons. Oneof these was the struggle participants had with managing clear and non-clear water.This management was made more difficult because a number of them felt hesitant tolet the floor get wet, because it would make a “water mess”. They therefore got soap andwater rinsed from their body back in the basin, mixing it with clear water. Designingthe space around the basin and seat explicitly as a wet space will make it easier tokeep the water in the basin clean.

Finally, an issue that seems to be crucial for splashing to work or not as a lessresource intensive alternative to showering is the issue of getting cold. Even thoughparticipants were fully dressed, did not use actual water and were in fact instructednot to feel cold, some of them still complained about feeling cold during splashing. Inshowering, the body stays warm thanks to a constant supply of warm water, somethingthat exactly lacks in splashing. Whether participants would have indeed felt cold hadtheir performances included water and nakedness is something that cannot be knownfrom the current set-up. There is however, strong reason to expect that getting coldcould be a major issue. Therefore, one of the requirements for the redesign is to includesome form of heating.



Fig. 5. Design and images by Linus Knupfer for LivingGreen.org [Knupfer 2011]. Images with permissionof Linus Knupfer.

5.4. The Refined Design and Next Steps

Based on the results of the study, a redesign was made of the ‘stuff for splashing’, bya designer who participated in the study and was elaborately briefed on the outcomes.In addition, he did experiments in his own bathroom to try out for example differentsizes for the basin.

The resulting design is shown in Figure 5. As can be seen, it is clearly different fromthe previous design shown in Figure 2. The basin is reduced to approximately 2 liters, ahandheld push-button shower is added and the fixture is designed for easy instalmentin existing shower cubicles. A local and quick heat source is added in the cubicle.

Besides the physical design, the redesign iteration also resulted in a set of instruc-tions and new terminology to explain possible ways of splashing to novices, includingparticular ways of positioning splashing relative to showering. Initial prototype testswith ten participants—this time with water—showed both positive responses to thewashing experience and, without explicit mention of the water saving objective, an av-erage water use of 20 liters. The water savings and with that the energy saved to heatthe water, more than make up for the energy required for the heater [Knupfer 2011].

Referring back to Hand et al. [2005], what also becomes clear from analysing the his-toric career of bathing, is that how change happens is not a matter of new technologicalpossibilities alone. The shower existed long before showering became popular. More-over, bathing does not exist in isolation, it is part of a field of interconnected practices[Schatzki 2001]. Ideas of cleanliness or perceptions of the body for example, are notelements of bathing alone. Reconfiguration of bathing may require reconfiguration ofa wide range of related practices. Using the new prototype, next studies will thereforeinvolve long term testing in real life environments to gain insight into the interactionof splashing with other domestic practices, such as sleeping, working, having breakfastand laundering.

6. REFLECTIONS ON THE APPROACH

While the empirical aim of the study was to generate a practice prototype of splashingand to inform further design iterations, the methodological aim was to explore how theimplications of taking practices as a unit of design could be integrated into a generativemethod. The proposed Generative Improv Performances (GIP) approach provides a



possible answer to this question. The following section will reflect on the strengths andlimitations of GIP in relation to existing prototyping approaches in (sustainable) HCI.

6.1. Isolating the Practice

A particular strength of GIP is its ability to generate rich narratives of possible futuresthat strongly deviate from the currently normal. While there is value in taking proto-types to the field and doing performances in context, as promoted by several authors(e.g., Odom et al. [2012], Foverskov and Binder [2009], Iacucci et al. [2002]), this articleargues that the lab environment has certain advantages over the field. As opposedto studies conducted in the context of people’s own homes, the lab set-up created anot-so-everyday crisis of routine in which the “breaking” and “shifting” of structuresis deliberately triggered. By making not so readily available the familiar resourcesused in the recreation of day-to-day conduct and by suggesting particular alternatives,the set-up facilitated the integration of familiar and unfamiliar elements into novelconfigurations of the practice.

However, this isolation of practices from their everyday context can also be seen asa weakness of the approach. Besides missing potentially important links with othereveryday practices, GIP can only be applied to practices that can be isolated in such away. This excludes for example the study of the Local Energy Indicator by Pierce andPaulos [2012], or communal energy management [Boucher et al. 2012], which involvea range of practices that are dispersed in time and over the home. Addressing thislimitation, a possible avenue of further research may be the use of GIP in a LivingLab setting [Bakker et al. 2010], where the lab comprises an entire dwelling andperformances span several days or more.

Another form of isolation in this setup is isolation of participants from each other.This was more of a practical than methodological choice, since it would have been nearlyimpossible to get all participants together at the same time. Isolating participants iscontrary to the idea of practices as collective achievements in which interaction betweenpractitioners is crucial [Schatzki 2001]. An ideal set-up may have been more like theSituative and Participative Enactment of Scenarios method [Iacucci et al. 2002], whereparticipants perform in front of each other. However, this idea of building on othersmay reduce the level of variety that is reached with the current set-up. Whether andhow GIP could include participant interaction is another topic of further study.

6.2. Improv Actors

Seemingly unique about the approach is the involvement of trained improvisationactors. Although other types of participants can be involved, GIP assumes that impro-visation acting, and importantly the ability to act outside of the currently normal, is askill that takes time and practice to learn. The thought provoking idea of re-thinkingco-design as social drama, introduced by Foverskov and Binder [2010], shows manysimilarities to the ideas behind GIP. Their observation that the actors, who were nottrained actors before being involved in the project, needed preparation lasting severalhours spread over several days before they could make a credible performance, confirmsthe idea that improv acting is a skill requiring training.

Besides saving preparation time, the use of skilled improv actors also seems toshift when and by whom the ‘designing’ is done. In the User Enactments method forexample, designers first make a detailed script or scenario, which users are then askedto enact [Odom et al. 2012]. In the case analysed by Foverskov and Binder [2010] theprocess of making the scenario was more participatory, but most of the designing hadtaken place before the scenario was played out. When using skilled improv actors, theactual performance can become the heart of where designing happens, making theparticipants the main designers.



This approach is basically different from evaluative prototyping studies in whichemphasis is placed on evaluation, addressing the main question of whether somethingworks. In GIP, the main question is how something might work. GIP forms a startingpoint for further research and debate on how decisions on “desirable” change are made,should be made and by whom; a discussion that has been addressed in a similarway in the related field of backcasting and planning by Wangel [2012]. Rather thanconceptualising future users as rational actors that can be persuaded to choose the“right” behavior, as determined by the designer, and using prototyping to evaluatewhether this behavior and expected savings are achieved, GIP involves practitionersin generating reconfigurations that may not seem logical or rational to the designer,but that make sense to them.

Another function of prototypes common in HCI is using them as probes [Gaveret al. 1999; Hutchinson et al. 2003; Routarinne 2007]. Such probes are set out inthe field for various durations of time to generate discursive, inspirational responsesfrom participants. Although generative rather than evaluative in nature, probes dif-fer from GIP in two main ways. One, because GIP is mainly interested in (detailed)bodily performances. When focusing on discursive responses, findings tend to comprisethings participants think they might do. However, this can differ strongly from whatthey would actually do, simply because in performance of the unfamiliar, unexpectedpracticalities are bound to arise. Something that repeatedly happened in the splashperformances. Two, because GIP actively facilitates reconfiguration of practices beyondthe usual suspects. Like observed in the Experiments in Practice and Trigger-Productstudies, situating a study in the context of normal daily life limits the level of changethat is seen as feasible. Although challenged to integrate novel elements (the probe),when people are not actively facilitated to reconfigure practices, they tend to drawfrom a readily available set of skills and images. GIP was developed to overcome initialresistance to change by actively challenging participants to make something that mayseem unacceptable at first sight (“You want me to wash from a bucket?”), work throughskilled improvisation.

Surely the choice for improv actors also has limitations. Bathing is a practice in whicheveryone is expert, but there are examples where target audiences are more specific.For example, someone improvising the role of a parent while not having children them-selves, or the role of a plumber while not being familiar with the actual practices ofplumbers. Although not being familiar with existing conventions may be an advantagewhen aiming for novel configurations, a major limitation will be that the actors willnot have the appropriate library of (potential) skills to draw on and therefore may lackthe ability to generate reconfigurations that make sense beyond the specific situationof the performance. Further study into the practices of improv acting could indicatewhether particular improv skills could be conveyed to a wider group of potential GIPparticipants.

7. CONCLUSIONS

By deducing guidelines from sociological theories of practice, this article has arguedthat taking practices as a unit of design means to generate reconfigurations of images,skills and stuff and their links, and requires the inclusion of bodily performance, thecreation of crises of routine and a variety of performances. The proposed GenerativeImprov Performances (GIP) approach integrates these guidelines, entailing a stage inthe form of a lab environment with rough prototypes, a scene directed by a set of looseconstraints, and improvisation actors as players.

In an empirical case study on bathing, GIP has shown to be capable of generatingreconfigurations of elements at a level that resonates with the scale of change that canbe observed from tracing practices back in time. Initial resistance to radical change



that is grounded in existing material and social structures can be overcome by involv-ing participants particularly skilled at making challenging situations work throughimprovisation.

By making bodily performance central to the approach, the locus of design activity isshifted from professional designers to participants. Moreover, by performing in detail,the gap is bridged between what people think they might do and what actually worksin an unfamiliar situation—a gap arguably even larger than the widely recognized gapbetween what people say they do and what they actually do in everyday routine.

The resulting range of performances flesh out opportunities for change that encom-pass but go beyond technological and behavioral change, showing potential for largeand enduring reductions in household resource consumption.

Although the empirical study did not entail an interactive system in its common def-inition, examples offered in the article indicate that the approach is equally applicablein the area of sustainable HCI and the discussion has revealed several avenues forfurther research.

ACKNOWLEDGMENTS

We want to thank all participants of the study for their time and valuable inputs, the two design graduatestudents for designing and making the prototypes used in the studies, reviewers of earlier versions of thisarticle for valuable and encouraging comments, and the editors of the special issue for their initiative anddedication.

REFERENCES

BAKKER, C., EIJK, D. V., SILVESTER, S., REITENBACH, M., JONG, A. D., AND KEYSON, D. 2010. Understandingand modelling user behaviour in relation to sustainable innovations: the LIVING LAB method. InProceedings of the International Symposium on Tools and Methods of Competitive Engineering. I.Horvath, F. Mardorli, and Z. Rusak, Eds., Delft University of Technology, Delft.

BOUCHER, A., CAMERON, D., AND JARVIS, N. 2012. Power to the people: Dynamic energy management throughcommunal cooperation. In Proceedings of the ACM Conference on Designing Interactive Systems. ACM,612–620.

BOURDIEU, P. 1977. Outline of a Theory of Practice. Cambridge University Press.BRYNJARSDOTTIR, H., HAKANSSON, M., PIERCE, J., BAUMER, E., DISALVO, C., AND SENGERS, P. 2012. Sustainably

unpersuaded: How persuasion narrows our vision of sustainability. In Proceedings of the SIGCHIConference on Human Factors in Computing Systems. ACM, 947–956.

BUSHMAN, R. AND BUSHMAN, C. 1988. The early history of cleanliness in America. J. Amer. History 74,1213–1238.

DAM VAN, S. S., BAKKER, C. A., AND VAN HAL, J. D. M. 2010. Home energy monitors: impact over themedium-term. Build. Res. Inf. 38, 458–469.

DE BORJA, J., KUIJER, L., AND APRILE, W. 2010. Sustainable systems innovations in households’ food acquisitionactivities: a practice oriented approach. In Proceedings of the ERSCP-EMSU. J. Q. R. Wever, A. Tukker,J. Woudstra, F. Boons, and N. Beute, Eds., Faculty of Industrial Design Engineering.

DISALVO, C., SENGERS, P., AND BRYNJARSDOTTIR, H. 2010. Mapping the landscape of sustainable HCI. InProceedings of the SIGCHI Conference on Human Factors in Computing Systems. ACM.

DOURISH, P. 2010. HCI and environmental sustainability: The politics of design and the design of politics. InProceedings of the ACM Conference on Designing Interactive Systems. ACM, 1–10.

Dutch Open Air Museum. Collection, Personal Care, Zo Fris Als Een Hoentje, http://www.openluchtmuseum.nl/index.php?pid=372&sub=29.

FEATHERSTONE, M. 1991. The Body: Social Process and Cultural Theory. Sage, London.FOEKEMA, H. AND THIEL, L. V. 2011. Watergebruik thuis 2010 TNS NIPO, in opdracht van VEWIN, Amsterdam.FOVERSKOV, M. AND BINDER, T. 2010. Design as everyday theatre. In Rehearsing the Future, J. Halse, E.

Brandt, B. Clark, and T. Binder, Eds., Danish Design School Press, Copenhagen, 204–210.FOVERSKOV, M. AND BINDER, T. 2009. Rehearsing the Future: in and out of Scenarios in a Reflective Practicum.

In Proceedings of the Nordic Design Research Conference. Oslo School of Architecture and Design.FROEHLICH, J., FINDLATER, L., AND LANDAY, J. 2010. The design of eco-feedback technology. In Proceedings of

the SIGCHI Conference on Human Factors in Computing Systems. 1999–2008.



GAVER, W. W., DUNNE, T., AND PACENTI, E. 1999. Design: Cultural Probes. ACM interactions 6, 21–29.GIDDENS, A. 1984. The Constitution of Society: Outline of the Theory of Structuration. Polity Press, Cambridge.GRAM-HANSSEN, K. 2008. Consuming technologies - developing routines. J. Cleaner Prod. 16, 1181–1189.GRAM-HANSSEN, K. 2009. Standby consumption in households analyzed with a practice theory approach. J.

Ind. Ecol. 14, 150–165.GRAM-HANSSEN, K. 2011. Understanding change and continuity in residential energy consumption. J.

Consumer Culture 11, 61–78.HAND, M., SHOVE, E., AND SOUTHERTON, D. 2005. Explaining showering: A discussion of the material,

conventional, and temporal dimensions of practice. Sociol. Res. Online 10.HIELSCHER, S., FISCHER, T., AND COOPER, T. 2008. The return of the beehives, Brylcreem and botanical!

An historical review of hair care practices with a view to opportunities for sustainable design. InProceedings of the Undisciplined! Design Research Society Conference. Design Research Society.

HUTCHINSON, H., WEND, M., ET AL. 2003. Technology probes: inspiring design for and with families. InProceedings of the SIGCHI Conference on Human Factors in Computing Systems. ACM, 17–24.

IACUCCI, G. AND KUUTTI, K. 2002. Everyday life as a stage in creating and performing scenarios for wirelessdevices. Personal Ubiq. Comput. 6, 299–306.

INGRAM, J., SHOVE, E., AND WATSON, M. 2007. Products and practices: Selected concepts from science andtechnology studies and from social theories of consumption and practice. Design Issues 23, 3–16.

ISH. 2009, 2011. The world’s leading trade fair. The Bathroom Experience, Building, Energy, Air-conditioningTechnology, Renewable Energies, Frankfurt, Germany.

JULIER, G. 2007. Design practice within a theory of practice. Design Principles Pract: Int. J. 1.KAPPEL, K. AND GRECHENIG, T. 2009. “Show-me”: WATER consumption at a glance to promote water conserva-

tion in the shower. In Proceedings of the 4th International Conference on Persuasive Technology. ACM,1–6.

KARAKAT, H. 2009. Designing for alternative (sustainable) bathing practices. In Industrial Design Engineer-ing, Delft University of Technology, Delft.

KNUPFER, L. 2011. An alternative bathing practice for sustainable renovation. In Industrial DesignEngineering, Delft University of Technology, Delft.

KORKMAN, O. 2006. Customer Value Formation In Practice: A Practice-Theoretical Approach. Departmentof Marketing and Corporate Geography, Swedish School of Economics and Business Administration,Helsinki, 196.

KUIJER, L. AND DE JONG, A. M. 2011. Practice theory and human-centered design: A sustainable bathingexample. In Proceedings of the Nordic Design Research Conference. T. H. Ilpo Koskinen, R. Maze, B.Matthews, and J.-J. Lee, Eds., School of Art and Design, Aalto University, Helsinki, Finland, 221–227.

KUIJER, L. AND DE JONG, A. 2012. Identifying design opportunities for reduced household resourceconsumption: Exploring practices of thermal comfort. J. Design Res. 10, 67–85.

LASCHKE, M., HASSENZAHL, M., DIEFENBACH, S., AND TIPPKAMPER, M. 2011. With a little help from a friend:A shower calendar to save water. In Proceedings of the SIGCHI Conference on Human Factors inComputing Systems. ACM, 633–646.

LATOUR, B. 1992. Where are the missing masses? The sociology of a few mundane artifacts. In ShapingTechnology/Building Society: Studies in Sociotechnical Change. W. E. Bijker and J. Law, Eds., MITPress, Cambridge, 225–258.

LATOUR, B. 1993. We Have Never Been Modern. Harvard University Press.MANZINI, E. 2009. Viewpoint: new design knowledge. Design Studies 30, 4–12.MATSUHASHI, N., KUIJER, L., AND JONG, A. D. 2009. A Culture-Inspired Approach to Gaining Insights for De-

signing Sustainable Practices. In Proceedings of the 6th International Symposium on EnvironmentallyConscious Design and Inverse Manufacturing. Japan Society of Mechanical Engineers, 547–552.

MEDLER, B. AND MAGERKO, B. 2010. The implications of improvisational acting and role-playing on designmethodologies. In Proceedings of the 28th International Conference on Human Factors in ComputingSystems. ACM, 483–492.

Merriam-Webster online dictionary. Definition of ‘bath’. http://www.merriamwebster.com/dictionary/bathing(Last accessed 11/9/12).

MUNNECKE, M. 2007. Future practices: Coshaping everyday life. In Proceedings of the ICSID & IDSA WorldDesign Congress.

ODOM, W., ZIMMERMAN, J., DAVIDOFF, S., FORLIZZI, J., DEY, A. K., AND LEE, M. K. 2012. A fieldwork of the futurewith user enactments. In Proceedings of the ACM Conference on Designing Interactive Systems. ACM,338–347.



OVERBEEKE, P. V. 2001. Kachels, Geisers en Fornhuizen; Keuzeprocessen en energieverbruik In Nederlandsehuishoudens 1920–1975. Uitgeverij Verloren, Hilversum.

PETTERSEN, I. N. 2009. Framing the role of technology in transformation of consumption practices: Beyonduser-product interaction. In Proceedings of the Conference on Joint Actions on Climate Change.

PIERCE, J. AND PAULOS, E. 2010. Materializing Energy. In Proceedings of the ACM Conference on DesigningInteractive Systems. ACM.

PIERCE, J. AND PAULOS, E. 2011. Second-hand interactions: investigating reacquisition and dispossessionpractices around domestic objects. In Proceedings of the Annual Conference on Human Factors inComputing Systems. ACM, 2385–2394.

PIERCE, J. AND PAULOS, E. 2012. The local energy indicator: designing for wind and solar energy systems inthe home. In Proceedings of the ACM Conference on Designing Interactive Systems. ACM, 631–634

RAVANDI, M., MOK, J., AND CHIGNELL, M. 2009. Development of an Emotional Interface for Sustainable WaterConsumption in the Home. In Proceedings of the 13th International Conference on Human-ComputerInteraction. Part III: Ubiquitous and Intelligent Interaction. Springer, 660–669.

RECKWITZ, A. 2002. Toward a theory of social practices: A development in culturalist theorizing. Eur. J. SocialTheory 5, 243–263.

RØPKE, I. 2009. Theories of practice: New inspiration for ecological economic studies on consumption. Ecol.Econ. 68, 2490–2497.

ROUTARINNE, S. 2007. Domestication as a design intervention. In Proceedings of the 2nd Nordic DesignResearch Conference.

SCHATZKI, T. R. 1996. Social Practices: A Wittgensteinian Approach to Human Activity and the Social.Cambridge University Press, Cambridge, UK.

SCHATZKI, T. R., CETINA, C. K., AND SAVIGNY, E. V. 2001. The Practice Turn in Contemporary Theory. Routledge,London.

SCOTT, K., BAKKER, C., AND QUIST, J. 2011. Designing change by living change. Design Studies 33, 279–297.SCOTT, K., QUIST, J., AND BAKKER, C. 2009. Co-design, social practices and sustainable innovation: Involving

users in a living lab exploratory study on bathing. In Proceedings of the Conference on Joint Actions onClimate Change.

SHOVE, E., PANTZAR, M., AND WATSON, M. 2012. The Dynamics of Social Practice: Everyday Life and How ItChanges. Sage, London.

SHOVE, E. AND PANTZAR, M. 2005. Consumers, Producers and Practices: Understanding the invention andreinvention of Nordic Walking. J. Consumer Culture 5, 34–64.

SHOVE, E., WATSON, M., HAND, M., AND INGRAM, J. 2007. The Design of Everyday Life. Berg.SHOVE, E. 2003. Comfort Cleanliness and Convenience: the Social Organisation of Normality. Berg, Oxford,

UK.STRENGERS, Y. A. A. 2011. Designing eco-feedback systems for everyday life. In Proceedings of the Annual

Conference on Human Factors in Computing Systems. ACM.VERA, D. AND CROSSAN, M. 2004. Theatrical improvisation: Lessons for organizations. Organiz. Studies 25,

727–749.WANGEL, J. 2012. Making Future: On Targets, Measures & Governance in Backcasting and Planning. KTH

Royal Institute of Technology, Stockholm.WARDE, A. 2005. Consumption and theories of practice. J. Consumer Culture 5, 131–153.WOODRUFF, A., HASBROUCK, J., AND AUGUSTIN, S. 2008. A bright green perspective on sustainable choices. In

Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. ACM, 313–322.ZIMMERMAN, J., STOLTERMAN, E., AND FORLIZZI, J. 2010. An Analysis and Critique of Research through Design:

towards a formalization of a research approach. In Proceedings of the ACM Conference on DesigningInteractive Systems. ACM, 310–319.

Received April 2012; revised January 2013; accepted June 2013


Documents

Practices as a Unit of Design: An Exploration of ...1043363/FULLTEXT01.pdf · of practice, it proposes that practice-oriented design ... HCI, improvisation actors, performances, practice